Nikaia
Menu
v0.0.1343pre-alpha GitHub

agent: builder model: opus max_turns: 120 budget_usd: 30 —

Brief: a failure inside members[S] or list(...) says the value’s path, at $.contacts[1].mail (ADR-grammar-2 D3, Part II 10.8, #1019 step 3d)

Facts: Traps already met

Score: #1019 step 3d, the error form of the typed read (ADR-344 D3’s $.contacts[3].email); S=A, O=~1.5M tokens, C=~1.5M tokens; grammar emitter and nikaia_std::grammar. Follows j4 (members[S], 20a79c9); does not need j5 and touches none of its files (j5: checker, comptime_*, build_values, dump; you: emit_grammar.nika, emit_pattern.nika, grammar.rs). If j5 is still running beside you, emit_*.nika may hold its uncommitted edits: git status first, and stage by name.

Problem (probe it with target/fast/nikaia run at your start, in your scratchpad)

struct Mail { addr: String, }
struct Contact { name: String, mail: Mail, }
struct Book { owner: String, contacts: Vec[Contact], }
grammar J {
	entry rule read[T] -> T = v:Value[T] eof -> { v }
	rule Ch -> i64 = x:any if(x != '"') -> { 1 }
	rule Key -> String = "\"" s:text(Ch*) "\"" ":" -> { s }
	rule Value[T] -> T = match T::kind {
		I64 => d:dec[i64](digit+) -> { d }
		Text => "\"" s:text(Ch*) "\"" -> { s }
		List(A) => "[" xs:list(Value[A], ",") "]" -> { xs }
		Struct(_) => v:Object[T] -> { v }
	}
	rule Object[S: Struct] -> S =
		"{" => m:members[S](Key, Value[_], ",") "}" -> { m }
}
fn main() {
	let b: Book = J::read[Book]("{\"owner\":\"a\",\"contacts\":[{\"name\":\"x\",\"mail\":{\"addr\":\"m\"}},{\"name\":\"y\",\"mail\":{\"adr\":\"m\"}}]}") catch {
		println(f"{error}")
		return
	}
	println(b.owner)
}

Today (architect’s probe on 20a79c9): the headline the member adr names no field at line 1, column 85, the excerpt, then 21 in lines, 13 of them the compiler’s own helpers (in __nikaia_members_2_one, _all, _more, _value, …), and no path. For step 2, J::read[Vec[i64]]("[1,2,x]") (rules I64 and List(A) of the probe) prints today expected a digit; found unexpected token x at line 1, column 6, the excerpt, then in item 1, in Value__i64, in item 2, in Value__Vec_i64_, in read__Vec_i64_: the in item N lines are the backend’s own frames (find what they count before relying on them; they are not the list index 2). (An NK1111 warning on the literal’s }} comes first and its help: drops a brace: #1091, not yours.)

Decision (architect, 2026-10-31, Language decisions: obvious; Part II 10.8 already says “the field names and list indices on its stack are the value’s path”)

Locations

To build

  1. A failure inside members[S] says the path of fields, and the helpers’ frames are gone. Test tests/compiler/tests/grammar_members_the_error_names_the_path.nika + .stdout: the probe above printing the whole error, plus inputs for a value that fails inside a member ("name":5 in contact 0: at $.contacts[0].name), a missing field one struct down (at $.contacts[0].mail or the struct the field belongs to), and a top-level unknown member (at $).
  2. A failure inside list(...) in a typed rule says the index. Test tests/compiler/tests/grammar_list_the_error_names_the_index.nika + .stdout: J::read[Vec[i64]]("[1,2,x]") with a grammar like the probe’s, printed whole: at $[2].
  3. An existing test whose expected output listed in __nikaia_* lines is updated (name it in Body:); grammar_failing_action_in_alternative and every other grammar_* test without members/list print exactly as before. One commit for all is fine; it says Part: #1019.

Done when

The new tests pass through nikaia test (one program alone: docs/builder-facts.md, Traps already met) with --both-settings, each failed once on the compiler without your change (say so under Tests:); all *grammar* tests pass; lower-std fixpoint clean; the cheap checks of Rules. Changelog: one line.

Stop and report if

the path cannot be carried without a change to winnow-grammar (say which function and what it must carry, Needs: architect decision); an edit outside Locations and the new tests; an error in another agent’s file. Stop about 15 turns before max_turns, commit what is finished and report partial.

Run 2 (architect, 2026-10-31; queued only after j8’s winnow-grammar commit is pushed upstream)

Run 1 (j6-1019-error-path.md.run1.report.md) stopped rightly: the “no backend change” route of Locations is wrong, a Cut error lost its item N frame. Decided (obvious): j8 makes winnow-grammar number a Cut error in a repetition as it numbers a Backtrack one; nothing else changes in the backend. What changed for you:

Rules (CLAUDE.md; this block is pasted into every brief, also a fix-forward)

Report (at most 15 lines, to exactly /home/user/Nikaia/docs/briefs/tools/j6-1019-error-path.md.report.md)

Fields of docs/architect.md, Briefs and reports: Result: done|noop|partial|blocked (the bare word alone on that line: Result: done, never a summary after it, which cuts the batch, b869a 2026-10-27; Owner: is a question for the owner or none, never a role), Files, Commit, Body (optional, one paragraph for the commit message), Fixes, Pushed, Tests, Gate (green or red <check>: your cheap checks, never none, which cuts the batch, j4 2026-10-31), Deviation, Found, Needs, Changelog, Owner, Disk freed, and for a move Rs2nika: (the kept N of M lines of rs2nika --score, then the three costliest hand edits; none for a task that moves no Rust, none (redesign) for a brief with Shape: redesign and no Convert: line). none where nothing, and then the bare word: Deviation: none (a note) cuts the batch (batch.nika compares exactly); a note goes into the commit message. Found: is only for a defect outside your change (it cuts the batch); a process event such as another session resetting or rewriting your files is no Found:, it goes under Needs: (m3a 2026-10-28); a converter draft that came out wrong is no Found:, it is one of the costliest edits under Rs2nika:, which the converter lane reads (e12-4b 2026-10-27); how you solved your own step goes into the commit message. Deviation: is only a step of the brief built otherwise than written or not built (a choice the brief left to you, such as keep what X needs and say so, is no deviation: it goes into the commit message; nor is building two steps in one commit when every step is built, e12-3a 2026-10-27, nor a step another commit on develop already built, shown by your test failing with that commit’s change reverted, b889 2026-10-27: name the commit in Body:, nor a test’s expected .stdout/.stderr taken from what the program really prints where the brief’s sketch of it was short or wrong, b934 2026-10-29: say so in Body:, nor a lowering that follows the ADR governing it where the brief’s expected lowering contradicts that ADR, b933 2026-10-29: name the ADR in Body:, nor a lowering other than the brief’s where rustc cannot take the brief’s, b1029 2026-10-31: say why in Body:, nor re-keying an allow-list entry (KNOWN) whose hash changed for the same program and the same reason, b-fixforward-sweep2 2026-10-28: say so in Body:, nor an existing test’s expected .stdout/.stderr updated because your change rightly changes the message it prints, b1012b 2026-10-31: name the test in Body:); how you verified (e.g. a test not run against the unfixed compiler) goes under Tests:. A noop says why under Tests: (what was probed), not under Deviation:. Final answer, one line: done <sha>, report <path> (or BLOCKED: ... / PUSH REJECTED ...).

Attribution: Co-Authored-By: Claude Opus 5.5 noreply@anthropic.com Claude-Session: https://claude.ai/code/session_01UVxJDvLe1uS47UcqA9y1U2

This site is open source. Improve this page.